From the Wired article:

Initially, the equations generated by the program failed to explain the data, but some failures were slightly less wrong than others. Using a genetic algorithm, the program modified the most promising failures, tested them again, chose the best, and repeated the process until a set of equations evolved to describe the systems. Turns out, some of these equations were very familiar: the law of conservation of momentum, and Newton’s second law of motion.

I’m very happy with this news article, I had a similiar idea some time ago, but little different, I’m currently developing it. While in development I always asked myself if it’ll works, but with this good news, maybe it works, but I’ve many work to do yet =)

Read the full article

UPDATE 08/04: you can read the supplemental materials and the report on the Cornell CCSL site.